Machine learning prediction of formation enthalpy of multicomponent sigma phase

The dataset contains the predicted formation enthalpy of intermetallic sigma phase from uniary up to quinary system. The formation enthalpy can be used to calculate the stability, site occupancy and be used to prepare TDB files for phase diagram calculations. The dataset includes the Gibbs energy calculated for equimolar composition of sigma phase using the compound energy formalism. A script is attached to calculate the Gibbs energy of sigma phase using the compound energy formalism. Site occupancy can be extracted. In addition, the data set includes the all the training data/test data, obtained from DFT calculation, that were used to train the graph neural network (ALIGNN).